I50 BIRD LIFE GLIMPSES 



capable of producing the effect in question. The 

 starlings then — this, at least, is my own conclusion 

 — though they seem to fly all together, in one long 

 string, really do so in regiment after regiment, and, 

 moreover, there is a certain order — and that a 

 strange one — by which these regiments leave the 

 plantation. It is not the first ones — those, that is 

 to say, that are stationed nearest the dormitory — 

 that lead the flight out to it, but the farthest or 

 back regiments, rise first, and fly, successively, over 

 the heads of those in front of them. Thus the 

 plantation is emptied from the farther end, and that 

 part of the army which was, in sitting, the rear, 

 becomes, in flying, the van. This, at least, seems to 

 be the rule or tendency, and precisely the same 

 thing is observable with rooks, though in both it 

 may be partially broken, and thus obscured. One 

 must not, in the collective movements of birds, 

 expect the precision and uniformity of drilled human 

 armies. It is, rather, the blurred image, or confused 

 approximation towards this, that is observable, and 

 this is, perhaps, still more interesting. 



One more point — and here, again, rooks and star- 

 lings closely resemble each other. It might be 

 supposed that birds thus flying, in the dusk of 

 evening, to their resting-place, would be anxious to 

 get there, and that the last thing to occur to them 

 would be to turn round and fly in the opposite 

 direction. Both here, however, and in the flights 

 out in the morning, we have that curious phenome- 

 non of breaking back, which, in its more salient 

 manifestations, at least, is a truly marvellous thing 

 to behold. With a sudden whirr of wings, the 



